Metcalfe's a good guy...
... but his opinion on the subject of MOOCs is hardly more valuable than that of J Random Idiot. As far as I'm aware, he hasn't conducted any pedagogical research. He probably doesn't even have much in the way of pedagogical training; many university professors don't, particularly in scientific and technical fields (Metcalfe is a prof of EE).
MOOCs are swell for many purposes - they're particularly nice for people looking to achieve some reliable knowledge about a subject area where they don't already have expertise, or to give someone the flavor of a field they're thinking about trying, or just for general broadening of the mind. I would never discourage someone from taking one. But they are by no means well-suited to every subject, or to every student.
They work well for students who are reasonably self-motivated and not easily discouraged when learning in a traditional instructional environment, and for subjects where the knowledge-transfer model works well. Process-model subjects like composition are much harder to adapt effectively into the MOOC model.
(The less said about that twee "BOOC" metaphor, the better.)
And let's not forget that his track record with predictions is not great.
And by the way, Rik, MOOC is an acronym for Massive Open Online Course. I don't know what "massively open" might entail ("dude, that course is massively open!"), but the intent of the phrase is that "massive" modifies "course", just like the other two adjectives.